InferenceBench

llm.mt.flores-200-mini-en-fr

5 entries. Pareto frontier computed on throughput_tok_per_s (higher is better) vs. ttft_p50_ms (lower is better). Rows marked P are on the frontier.

5 of 5 matching
Model Engine Hardware Quant TTFT P50 (ms) TTFT P99 (ms) Throughput (tok/s) $/M tokens J/token Power avg (W) Power peak (W) WER mean J / audio s Envelope
meta-llama/Llama-3.1-70B-Instruct vllm unknown 8x NVIDIA H100 80GB HBM3 fp16 32.03 JSON
mistralai/Mistral-7B-Instruct-v0.3 vllm unknown 8x NVIDIA H100 80GB HBM3 fp16 12.00 JSON
Qwen/Qwen2.5-7B-Instruct vllm unknown 8x NVIDIA H100 80GB HBM3 fp16 14.11 JSON
google/gemma-2-9b-it vllm unknown 8x NVIDIA H100 80GB HBM3 fp16 16.77 JSON
meta-llama/Llama-3.1-8B-Instruct vllm unknown 8x NVIDIA H100 80GB HBM3 fp16 14.16 JSON